• Bourke, W., McAvaney B. , Pury K. , and Thurling R. , 1977: Global modeling of atmospheric flow by spectral methods. Methods in Computational Physics, Vol. 17, J. Chang, Ed., Academic Press, 267–324.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., 1991: Mid-latitude Weather Systems. Harper Collins, 507 pp.

  • Davies, R., 1982: Documentation of the solar radiation parameterization in the GLAS climate model. NASA Tech. Memo. 83961, 57 pp.

  • DeWitt, D. G., 1996: The effect of the cumulus convection scheme on the climate of the COLA general circulation model. COLA Tech. Rep. 27, 58 pp. [Available from COLA, 4041 Powder Mill Rd., Suite 302, Calverton, MD 20705.].

    • Search Google Scholar
    • Export Citation
  • Dirmeyer, P. A., and Zeng F. J. , 1999: An update to the distribution and treatment of vegetation and soil properties in SSiB. COLA Tech. Rep. 78, 25 pp. [Available from COLA, 4041 Powder Mill Rd., Suite 302, Calverton, MD 20705.].

    • Search Google Scholar
    • Export Citation
  • Fennessy, M. J., and Coauthors. 1994: The Simulated Indian Monsoon: A GCM sensitivity study. J. Climate, 7 , 3343.

  • Harshvardhan, R., Davies R. , Randall D. A. , and Corsetti T. G. , 1987: A fast radiation parametrization for atmospheric circulation models. J. Geophys. Res., 92 ((D1),) 10091016.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hastenrath, S., 1985: Climate and Circulation of the Tropics. Kluwer Academic, 455 pp.

  • Kasahara, A., 1974: Various vertical coordinate systems used for numerical weather prediction. Mon. Wea. Rev., 102 , 509522.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kiehl, J. T., Hack J. J. , Bonan G. , Boville B. A. , Williamson D. L. , and Rasch P. J. , 1998: The National Center for Atmospheric Research Community Climate Model: CCM3. J. Climate, 11 , 11311149.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kinter, J. L., Shukla J. , Marx L. , and Schneider E. K. , 1988: A simulation of the winter and summer circulations with the NMC global circulation model. J. Atmos. Sci., 45 , 24862522.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kinter, J. L., and Coauthors. 1997: The COLA atmosphere–biosphere general circulation model. Volume 1: Formulation. COLA Tech. Rep. 51, 1–44. [Available from COLA, 4041 Powder Mill Rd., Suite 302, Calverton, MD 20705.].

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Suarez M. J. , 1995: Relative contributions of land and ocean processes to precipitation variability. J. Geophys. Res., 100 , 1377513790.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Heiser M. , 2000: Variance and predictability of precipitation at seasonal-to-interannual timescales. J. Hydrometeor., 1 , 2646.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lacis, A. A., and Hansen J. E. , 1974: A parameterization for the absorption of solar radiation in the earth's atmosphere. J. Atmos. Sci., 31 , 118133.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mellor, G. L., and Yamada T. , 1982: Development of a turbulence closure model for geophysical fluid processes. Rev. Geophys. Space Phys., 20 , 851875.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miyakoda, K., and Sirutis J. , 1977: Comparative integrations of global spectral models with various parametrized processes of subgrid scale vertical transports. Beitr. Phys. Atmos., 50 , 445447.

    • Search Google Scholar
    • Export Citation
  • Moorthi, S., and Suarez M. J. , 1992: Relaxed Arakawa–Schubert: A parameterization of moist convection for general circulation models. Mon. Wea. Rev., 120 , 9781002.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Palmer, T. N., Shutts G. J. , and Swinbank R. , 1986: Alleviation of a systematic westerly bias in general circulation and numerical weather prediction models through an orographic gravity wave parametrization. Quart. J. Roy. Meteor. Soc., 112 , 10011039.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Polcher, J., and Laval K. , 1994: A statistical study of the regional impact of deforestation on climate in the LMD GCM. Climate Dyn., 10 , 205219.

  • Reale, O., Dirmeyer P. , and Schlosser A. , 2002: Modeling the effect of land surface evaporation variability on precipitation variability. Part II: Time- and space-scale structure. J. Hydrometeor.,3, 451–466.

    • Search Google Scholar
    • Export Citation
  • Reynolds, R. W., and Smith T. M. , 1994: Improved global sea surface temperature analyses using optimal interpolation. J. Climate, 7 , 929948.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sellers, P. J., Mintz Y. , Sud Y. C. , and Dalcher A. , 1986: A Simple Biosphere Model (SiB) for use within general circulation models. J. Atmos. Sci., 43 , 505531.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Simmons, A. J., and Struefing R. , 1981: An energy and angular-momentum conserving finite-difference scheme, hybrid coordinates and medium-range weather prediction. ECMWF Tech. Rep. 28, 68 pp. [Available from ECMWF, Shinfield Park, Reading RG2 9AX, United Kingdom.].

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., Walker G. K. , Kim J-H. , Liston G. E. , Sellers P. J. , and Lau W. K-M. , 1996: Biogeophysical consequences of a tropical deforestation scenario: A GCM simulation study. J. Climate, 9 , 32253247.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tiedtke, M., 1984: The effect of penetrative cumulus convection on the large-scale flow in a general circulation model. Beitr. Phys. Atmos., 57 , 216239.

    • Search Google Scholar
    • Export Citation
  • Xue, Y-K., Sellers P. J. , Kinter J. L. , and Shukla J. , 1991: A simplified biosphere model for global climate studies. J. Climate, 4 , 345364.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xue, Y-K., Zeng F. J. , and Schlosser C. A. , 1996: SSiB and its sensitivity to soil properties. A case study using HAPEX–Mobilhy data. Global Planet. Change, 13 , 183194.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery
    Fig. 1.

    (a) Mean precipitation (mm day−1) for the control run. The mean is calculated from 48 yr of integration (1951–98). All the subsequent means are computed with the same time frame. (b) AL mean precipitation departure from the control. (c) AO mean precipitation departure from the control run. (d) Simulation A mean precipitation departure from the control run. The solid (dotted) line represents areas where the increase (decrease) in precipitation is statistically significant.

  • View in gallery
    Fig. 1.

    (Continued)

  • View in gallery
    Fig. 2.

    (a) Mean evaporation (mm day−1) for the control run. (b) AL mean evaporation departure from the control run. (c) AO mean evaporation departure from the control run. (d) Simulation A mean evaporation departure from the control run. The solid (dotted) line represents areas where the increase (decrease) in evaporation is statistically significant

  • View in gallery
    Fig. 2.

    (Continued)

  • View in gallery
    Fig. 3.

    (a) Mean surface temperature (°C) for the control run. (b) AO mean surface temperature departure from the control run

  • View in gallery
    Fig. 4.

    (a) Mean divergence at σ = 0.9205 (s−1) for the control run. (b) AL mean divergence departure from the control run. (c) AO mean divergence departure from the control run. (d) Simulation A mean divergence departure from the control run

  • View in gallery
    Fig. 4.

    (Continued)

  • View in gallery
    Fig. 5.

    (a) Mean divergence at σ = 0.275 (s−1) for the control run. (b) AL mean divergence departure from the control run. (c) AO mean divergence departure from the control run. (d) Simulation A mean divergence departure from the control run

  • View in gallery
    Fig. 5.

    (Continued)

  • View in gallery
    Fig. 6.

    (a) Mean relative vorticity at σ = 0.5 (s−1) for the control run. (b) AO mean vorticity departure from the control run. Cyclonic vorticity is light (dark) in the Northern (Southern) Hemisphere

  • View in gallery
    Fig. 7.

    (top) Mean vertical velocity (10−3 hPa s−1) between 22.5°W and 15°E: (left) control run, (middle) AO, and (right) AO departure from the control run. Contour intervals are every 0.01 × 10−3 hPa s−1 (control and AO) and every 0.0025 × 10−3 hPa s−1 (AO minus control), with the zero contour omitted. (bottom) Meridional wind (m s−1): (left) control run, (middle) AO, and (right) AO departure from the control run. Contour intervals are every m s−1 (control and AO) and every 0.25 s−1 (AO minus control), with the zero contour omitted

  • View in gallery
    Fig. 8.

    Mean precipitation (mm day−1) for (left) DJF and (right) JJA: (a), (b) control run; (c), (d) AL mean precipitation departure from the control run; (e), (f) AO mean precipitation departure from the control run; (g), (h) simulation A mean precipitation departure from the control run

  • View in gallery
    Fig. 9.

    Histogram with the variance of annual precipitation for each experiment: (a) tropical land (11.35°S–11.35°N); (b) Amazon basin (11.35°S–0°, 74.625°–50.625°W); (c) Congo basin (7.425°S–3.715°N, 13.15°–31.875°E); (d) Northern Hemisphere midlatitude land (29.685°–51.955°N); (e) continental United States to the east of the Rockies (29.685°–51.955°N, 110.625°–88.125°W); and (f) continental Eurasia (44.535°–63.085°N, 13.125°–136.875°E)

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 173 32 5
PDF Downloads 34 12 2

Modeling the Effect of Land Surface Evaporation Variability on Precipitation Variability. Part I: General Response

Oreste RealeCenter for Ocean–Land–Atmosphere Studies, Calverton, Maryland

Search for other papers by Oreste Reale in
Current site
Google Scholar
PubMed
Close
and
Paul DirmeyerCenter for Ocean–Land–Atmosphere Studies, Calverton, Maryland

Search for other papers by Paul Dirmeyer in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

This is the first of a two-part article that investigates the impact of land surface evaporation variability on the interannual variability of precipitation and compares it with the impact caused by sea surface temperature variability. Previous works by Koster and Suarez and Koster et al. provide the general strategy to control oceanic and land surface evaporation. For this part of the study, their numerical experiments are repeated using the Center for Ocean–Land–Atmosphere Studies (COLA) general circulation model. However, emphasis is put on the dynamics of the response, including a discussion of the changes in the mean climate; in particular, it is observed that the suppressed land evaporation variability changes the mean Northern Hemisphere storm track and the mean position of the intertropical convergence zone, which in turn affect the mean precipitation.

The analysis of the precipitation variance reveals a general agreement with previous works for the midlatitudes, whereas in the Tropics a stronger land-induced signal is detected. Furthermore, important regional differences in the response are found. Specifically, there is a predominant land signal over the Amazon region, in contrast to an equivalence between the land and ocean forcings over the Congo basin region. Finally, the model appears to be slightly more sensitive to seasonal–interannual variations of the land forcing than is the one adopted by Koster et al.

Current affiliation: Goddard Earth Sciences and Technology Center (UMBC)/Data Assimilation Office, Greenbelt, Maryland

Corresponding author address: Dr. Oreste Reale, Data Assimilation Office, Code 910.3, NASA Goddard Space Flight Center, Greenbelt, MD 20771. Email: oreale@dao.gsfc.nasa.gov

Abstract

This is the first of a two-part article that investigates the impact of land surface evaporation variability on the interannual variability of precipitation and compares it with the impact caused by sea surface temperature variability. Previous works by Koster and Suarez and Koster et al. provide the general strategy to control oceanic and land surface evaporation. For this part of the study, their numerical experiments are repeated using the Center for Ocean–Land–Atmosphere Studies (COLA) general circulation model. However, emphasis is put on the dynamics of the response, including a discussion of the changes in the mean climate; in particular, it is observed that the suppressed land evaporation variability changes the mean Northern Hemisphere storm track and the mean position of the intertropical convergence zone, which in turn affect the mean precipitation.

The analysis of the precipitation variance reveals a general agreement with previous works for the midlatitudes, whereas in the Tropics a stronger land-induced signal is detected. Furthermore, important regional differences in the response are found. Specifically, there is a predominant land signal over the Amazon region, in contrast to an equivalence between the land and ocean forcings over the Congo basin region. Finally, the model appears to be slightly more sensitive to seasonal–interannual variations of the land forcing than is the one adopted by Koster et al.

Current affiliation: Goddard Earth Sciences and Technology Center (UMBC)/Data Assimilation Office, Greenbelt, Maryland

Corresponding author address: Dr. Oreste Reale, Data Assimilation Office, Code 910.3, NASA Goddard Space Flight Center, Greenbelt, MD 20771. Email: oreale@dao.gsfc.nasa.gov

1. Introduction

The interannual variability of precipitation is the outcome of different processes: the internal dynamics of the atmosphere and the variability of boundary forcings. The boundary forcing considered here is evaporation from the ocean and the land. The purpose of this two-part study is to investigate the impact of marine and terrestrial evaporation variabilities on interannual precipitation variability. A set of general circulation model (GCM) experiments, in which evaporation variabilities from land and sea are in turn enabled or disabled, is performed. Previous works by Koster and Suarez (1995, hereafter KS95) and Koster et al. (2000, hereafter KA00) are the starting point for this article, and provide the general strategy adopted to control oceanic and land surface evaporation.

In KS95 the authors perform four 10-yr GCM simulations in which interannual evaporation variabilities are prescribed. In the control run the atmosphere (A), the land (L), and the ocean (O) vary, hence the name ALO. The state of the atmosphere and of the land surface variables is calculated by the model at each time step, whereas the state of the ocean surface varies according to a set of observed monthly sea surface temperatures (SSTs). To distinguish the control run from a different experiment in the second part of this study (Reale et al. 2002, hereafter Part II), we call it ALO1.

The other experiments are named according to the principle that when one of the two boundary sources of evaporation variability is suppressed, the corresponding letter (L or O) is dropped from the experiment name. For example, in the simulation named AL, the interannual ocean (O) variability is disabled by prescribing climatological, instead of interannually varying, sea surface temperatures. Only the atmosphere (A) and the land (L) are allowed to vary, hence the name AL. In the simulation AO, the land (L) evaporation variability is disabled by prescribing a set of monthly mean “evaporabilities” calculated from the control run. In simulation A, both ocean and land evaporation variabilities are prescribed. By doing so, the authors believe that they could estimate the interannual precipitation variability inherent to the atmospheric model dynamics. The comparison of the simulations in KS95 suggests that the ocean controls the timing of precipitation anomalies, whereas the land–atmosphere feedbacks can modulate their amplitude.

In KA00 the authors acknowledge the intrinsic limitations of KS95 and perform the same set of experiments, but with an improved model, longer integrations and with a large ensemble. The control run (ALO) and the AO run consist of a 16-member ensemble of 45-yr integrations; the AL and A runs consist of a 200-yr four-member ensemble. To these runs, the authors added a fifth simulation (ALOX), which allows terrestrial evaporation to vary on interannual scales but not on submonthly scales. The results suggest that the amplification of precipitation variance by land–atmosphere feedback is larger outside of the regions that are most affected by SSTs: namely, the midlatitude transitional zones. An important assumption of KA00 is a hypothesis of “linearity” and, therefore, “additivity” of the total variance of precipitation: the authors hypothesize that the contribution to total variance of the various components of the climate system can be separated or added together.

In the first part of this study the same set of experiments of KS95 and KA00 are performed with a different model. The main purpose is to verify if previous results can be reproduced and to what extent they are model dependent. Second, by analyzing the results on a more dynamical basis, we observe important changes in the mean general circulation of the atmosphere that arise by suppressing interannual land evaporability. Finally, given a general agreement with KA00 results, we notice that the overall response of the experiments needs to be analyzed carefully from a regional perspective, since important differences can be detected between different areas in the Tropics and in the midlatitudes. In Part II we address the validity of the hypothesis of linearity used in KA00 and, with the aid of three other experiments, decompose the land surface forcing on different timescales.

2. Model and experiment design

a. The model

The atmospheric model used in the Center for Ocean–Land–Atmosphere Studies (COLA) GCM version 2.2, a spectral model that can be considered a merging of the COLA GCM version 1, described by Kinter et al. (1988, 1997), with the National Center for Atmospheric Research (NCAR) Climate Community Model version 3 (CCM3), described in Kiehl (1998). As a general merging strategy, the dynamical core of CCM3 (Bourke et al. 1977) was installed into the COLA GCM, whereas the physics of the COLA model was retained. The main features of the COLA GCM version 2.2 are the following:

The atmospheric model is coupled to the so-called Simplified Simple Biosphere Model (SSiB), documented in Xue et al. (1991, 1996), and derived from the the Simple Biosphere Model (SiB; Sellers et al. 1986). However, the 2.2 version of the COLA GCM also contains an update of the vegetation and soil properties (Dirmeyer and Zeng 1999). In our experiments, a triangular truncation at wavenumber 31 is adopted, which corresponds to a horizontal resolution of approximately 3.75°.

b. The experiments

Four 49-yr experiments are performed, using the same strategy adopted in KA00. They are all named accordingly: ALO1 (control), AL, AO, and A. In experiment ALO1, the SST is from the 49-yr Reynolds and Smith (1994) updated dataset. Sea ice is also prescribed accordingly to the same dataset, whereas all the other land boundary forcings (snow cover, soil moisture, etc.) are calculated at each time step. The observed atmospheric initial conditions (inclusive of snow cover) are set on 1 February 1950, and they are obtained from the National Centers for Environmental Prediction (NCEP)–NCAR reanalyses dataset. The first 11 months of integration are discarded to eliminate problems of land surface spinup.

In experiment AL, the SSTs vary on a monthly climatology calculated from the same dataset. Thus, a control over the interannual variability of marine evaporation is exerted, and the ocean (O) variability is “disabled.”

Disabling the land (L) evaporation variability is conceptually more difficult, due to the intrinsic difference between SST's and soil moisture's control of evaporation and to the relatively poor observational knowledge of the soil moisture field. The concept of “evaporability” or “evaporation efficiency” is therefore adopted, following KS95 and KA00, in order to control continental evaporation in the runs in which L is to be disabled. The evaporation efficiency is defined as the ratio between actual evaporation
i1525-7541-3-4-433-e1
[where ρ is the air density, ε is constant, ps is the surface pressure, es is the saturated vapor pressure at the surface temperature, ea is the vapor pressure in the free air (viz., the lowest σ level), raer is the aerodynamic resistance, and rsurf is the surface resistance, computed as a parallel resistance of all the resistances that contribute to it] and potential evaporation Ep, the latter calculated from the former if the surface resistance rsurf is set to zero:
i1525-7541-3-4-433-e2
For each month m and each year y the ratio
i1525-7541-3-4-433-e3
is calculated throughout all time steps i of that particular m and y. This allows the definition of a time series of 48 × 12 monthly β from the control run ALO1, for each grid box. Averaging for each month m throughout the 48 yr, we obtain a set of 12 monthly βm, which represents our forcing. In a given month m, the actual evaporation E(i) for case AO, in which land is prescribed, is calculated by
EiEpiβm
at each time step. The βm used is the same for the same month m of different years, so it does not have interannual variability. In SSiB, a separate βm must be determined for soil evaporation and for canopy evapotranspiration, because the two elements of the land surface have distinct thermodynamic calculations. For vegetation, the rsurf term also includes stomatal resistance. It is important to note that, because of the variations in the atmospheric conditions, continental evaporation can still vary in the run with controlled land, exactly as marine evaporation can vary in the runs in which the SSTs are climatological.

Simulation A is performed with the same values of βm used in run AO, and with the same set of climatological SSTs adopted in run AL. So, the only source of evaporation variability in this run is assumed to be from the internal atmospheric dynamics.

3. Experiments results

a. Changes in the mean climate

1) Precipitation

In Fig. 1a the mean annual precipitation for the control run is displayed. A reasonable agreement with KA00 is found. However, one underlying assumption in the KA00 study is that, by perturbing the marine or terrestrial forcing, only the changes in precipitation variance are affected. The authors do not investigate changes in the mean climate consequent to the different forcings. In KS95, the authors show some change in the mean precipitation consequent to the different forcings (only in the Tropics), but this change seems to be reduced in KA00. However, investigation into the atmospheric dynamics leading to these changes is not performed in either of the two studies. In this work, changes in mean precipitation are examined, and found significant, also outside the tropical region. Moreover, changes in the mean temperature, divergence, and vorticity fields are investigated to understand how the changed boundary forcings propagate via the model's dynamics.

Figure 1b shows the change in mean annual precipitation that results from replacing the observed monthly varying SSTs with the monthly climatology (AL). The changes are significant mostly in the tropical Pacific, due to the suppression of the El Niño–Southern Oscillation (ENSO) signal. In fact, the AL experiment is essentially an experiment forced by “neutral” ENSO conditions. The typical precipitation signal “neutral minus El Niño” is characterized by a negative anomaly along the equator in the central Pacific, which is consistent with our results. However, we can observe other features that cannot be attributed to the ENSO suppression alone: for example, an apparent strengthening of the mean Indian monsoon seems to arise from the suppression of the interannual SST variability.

Figure 1c displays the effects on mean annual precipitation obtained by prescribing land evaporability (AO). An apparent weakening of the Asian monsoon is observed, together with a general, significant reduction of precipitation over land and an increase over ocean. This effect was also observed in the KS95 experiment. In our experiment, however, changes are not confined to the Tropics or to the marginal areas (the transitional areas between desert and moister climates, characterized by sharp precipitation gradients), but they appear in the midlatitudes as well. Particularly significant is the increase in precipitation along the northern Atlantic storm track, which is not observed in the KS95 experiment.

The response of the AL and AO experiments over the western Pacific, Indonesia, and India (Figs. 1b,c) are indicative of opposing effects of the oceanic and land forcing on the mean precipitation. This would suggest that once the ocean and land are altered together, their effects should somehow neutralize one another over these regions. Shown in Fig. 1d are the cumulated effects of removing both the ocean and the land forcing together. Over some areas, the net effect appears to be additive (i.e., equal to the sum of the AL and AO impacts). But over others, for example, India and Australia, the land impact dominates (i.e., no significant difference between experiments AO and A). In other words, at the regional level, the anomaly induced by land and ocean prescribed together (run A) is very different from the added anomalies of the runs in which land (AO) and ocean (AL) are prescribed separately. This is particularly evident for the Indian monsoon region. For instance, over the Bay of Bengal, there are negative values in the AL run and positive values in the AO run, of approximately the same magnitude. If additivity held true in the A run, values close to zero should be expected, but the values are again positive, and even larger than the AO run.

2) Evaporation

Figure 2 displays the mean evaporation in the control run (ALO1) and the corresponding anomalies. The AL minus control plot (Fig. 2b) shows that evaporation significantly increases in the eastern Pacific and decreases in the western Pacific, reflecting the suppression of the ENSO signal in a way even more evident than in Fig. 1b.

Figure 2c shows the evaporation of AO relative to the control simulation. An overall consistency with Fig. 1c is found (i.e., decrease of evaporation over land, where precipitation decreases, and increase of evaporation over ocean, where precipitation increases). However, the remarkable correspondence of large positive evaporation anomalies with the northern Atlantic and Pacific storm tracks suggests that the midlatitude baroclinic activity is somehow enhanced as a result of suppression of interannual variations of evaporabilities over land. This idea, that land can partly control midlatitude baroclinic activity (which is generally regarded as a pure product of the “internal atmospheric dynamics”), is quite intriguing and is discussed later in more detail. Here we are broadly defining “baroclinic activity” as the precipitation presumably generated by midlatitude weather systems, associated with horizontal temperature gradients, in order to separate it from the precipitation produced in the Tropics by the ITCZ and the monsoons.

Figure 2d shows the evaporation anomaly relative to experiment A. The same considerations applied to Fig. 1d can be applied here (regarding the nonadditivity of the contributions derived by ocean and land separately).

3) Surface temperature

Figure 3 displays the mean surface temperature in the control run (ALO1) and the corresponding AO minus control anomaly. Since the ocean SST is prescribed in all experiments, and the mean SST for the AL and A runs is calculated from the same SST dataset used to force the control, surface temperature variations in all experiments can occur only over land. However, the mean land surface temperature does not vary significantly as a consequence of using climatological SSTs (not shown). Conversely, a significant variation is observed in both runs (AO and A) in which the land properties are altered (no significant changes are present between the two, so only the AO case is shown). A general cooling over the Northern Hemisphere land is observed, particularly over the Sahara. In order to explain these temperature changes, the energy budget is computed at the surface for both runs:
i1525-7541-3-4-433-e5
where SW⇓ is downward shortwave radiation, SW⇑ is upward shortwave radiation, SH is sensible heat, NLW is net longwave radiation, and G is the net heat flux into the ground, considered as a residual. The total mean annual value of G is generally close to zero for both runs. However, the analysis of the various terms of the budget (not shown) show that there are regional and seasonal departures, which cause the observed temperatures changes. Over northern Africa, both upward and downward longwave radiation decrease, but the former prevails over the latter, so the net longwave from the surface decreases substantially, indicating lower surface temperatures. However, the cooler surface temperatures are caused by the suppression of the diurnal cycle of β. In fact, prescribing a mean fixed β throughout the month, without a diurnal cycle, can lead to enhanced evaporation during the day, which substantially reduces the daytime maximum temperatures. This cooling effect is also evident, more moderately, in the southern part of North America and Eurasia during summer, and is an artifact of the experiment design. It is indeed a limitation of this experiment, AO, and of KA00 as well. However, in Part II, another experiment is performed in which a mean monthly diurnal cycle of β is prescribed. The cooling over the Sahara disappears completely.

For the mid- and higher latitudes (i.e., mostly northern North America and northern Eurasia) the dominant term is an increase in upward shortwave radiation of approximately 5%–15% in the AO run, during the cold semester. For the same regions, there is a smaller decrease in latent heat and net longwave radiation during summer, whereas the other terms of the budget do not change significantly. The prevailing term however, responsible for the cooling, is SW⇑, and it is due to increased mean albedo caused by a longer duration of the snow on the ground. The persistence of the snow is consistent with lower surface temperatures.

The decrease in mean temperature over the northern part of the Northern Hemisphere continents leads to increased surface temperature gradients between ocean and land, particularly on the eastern sides of northern North America and Asia in the cold semester. This may be an indication of increased baroclinic activity and thus may contribute to an intensification of the storm track consistent with the increase in precipitation and evaporation observed in the AO run over the Northern Hemisphere storm tracks (Figs. 1c,2c).

4) Large-scale dynamics

To further investigate the impact of our changed boundary forcing on the atmospheric dynamics, and separate the effects involving convection from the effects involving midlatitude dynamics, we examine low-level convergence, upper-level divergence, and midtropospheric vorticity. In general, the following insights can be inferred by looking at these fields:

  • A simultaneous increase (decrease) in low-level convergence and upper-level divergence over some area where precipitation has increased (decreased) is an indication of an increase (decrease) in convective activity.

  • A midtropospheric vorticity anomaly associated with the tropical easterly jet is an indication of increased (decreased) horizontal shear, which may be related to increased (decreased) activity of easterly waves and, thus, precipitation (Hastenrath 1985), or to a shift in the intertropical convergence zone (ITCZ).

  • A mid-tropospheric vorticity anomaly, when associated with westerly flow, is an indication of changes in baroclinic instability (Carlson 1991), which predominantly affects midlatitude precipitation.

In Fig. 4a, the mean divergence at σ = 0.9205 (i.e., generally in the boundary layer) is displayed. The departure field relative to experiment AL is plotted in Fig. 4b and displays variations mostly over the tropical Pacific and over Indonesia, consistent with Figs. 1b and 2b. Figures 4c and 4d display the corresponding changes in the low-level divergence in experiments AO and A. Divergence fields typically show a great deal of small-scale structure, but the comparison with the upper-level divergence field allows us to delineate some features. The upper-level divergence for the control run and the corresponding experiments is plotted in Fig. 5. A comparison of precipitation and divergence changes in experiment AL reveals that the increase in precipitation over western Central America and over Indonesia and northern Australia (Fig. 1b) corresponds perfectly to an increase in low-level convergence (Fig. 4b) and upper-level divergence (Fig. 5b). Conversely, the decrease in precipitation observed over Indochina, the Philippines, and the western tropical Pacific (Fig. 1b) corresponds to a decrease in low-level convergence and upper-level divergence (Figs. 4b,5b). This means that the suppression of interannual SST variability in the AL experiment leads to a change in dynamics mostly confined to the Hadley cell and leaves the midlatitudes unaffected.

The interpretation of Figs. 1c, 4c, and 5c, relative to experiment AO, is not so straightforward, because the divergence field is noisier. However, opposite sign variations in low-level and upper-level divergence (indicative of changes in tropical convection) related to an increase in precipitation can be observed over a strip slightly north of the equator from India to the Philippines. Similarly, a relationship between a decrease in precipitation and a simultaneous decrease in low-level convergence and upper-level divergence can be observed over continental South Africa, Indonesia, and northern and eastern Australia. So, changes in convection related to the suppression of land variability are mostly confined to the tropical or subtropical land areas. These findings suggest that changes occurring outside the above-mentioned regions (mostly over the Northern Hemisphere) might be due to changes in baroclinic activity, as is discussed later. The comparison of Figs. 1d, 4d, and 5d leads to similar conclusions, although the effects of ocean and land are combined.

In Fig. 6a, the relative vorticity field at σ = 0.5, computed from the control run, is displayed. No significant change in the mean annual midtropospheric vorticity field is observed in the AL run (not shown). However, the seasonal anomalies reveal that the suppression of SST variability does increase the vorticity over the Northern Hemisphere in winter, particularly along the Atlantic storm track. This change, however, is smoothed out from the annual mean.

Conversely, the AO run displays a remarkable change in the annual mean with respect to the control run (Fig. 6b): a stationary wave pattern with two maxima and two minima between the Gulf of Guinea and Greenland. This wave pattern comprises two different dynamical responses. The maximum and minimum over the northern Atlantic would seem to indicate a strengthening of the westerly jet, related to an apparent increase in baroclinic activity (in agreement with Figs. 1c, 2c). However, this pattern results from an average across different structures arising in different seasons, as discussed further in the next section. The maximum and minimum over western Africa and the eastern tropical Atlantic suggests a southward displacement of the ITCZ. In fact, the African easterly flow, related to the ITCZ, generates cyclonic vorticity to the south and anticyclonic vorticity to the north of it. Thus, the vorticity and wind anomalies in Fig. 6b indicate a southward displacement and the strengthening of the African easterly jet.

This is confirmed by Fig. 7, in which the meridional cross section of vertical velocity and meridional wind, averaged between 22.5°W and 15°E, are displayed. The strengthening and southward displacement of the Hadley circulation is evident in the AO case. A similar pattern is seen in the mean vertical velocity and meridional wind calculated across longitudes spanning from 40°W to 40°E (not shown). The same section plotted for the AL case does not reveal substantial changes in the Hadley circulation over Africa and the tropical Atlantic, whereas there are important changes in the tropical Pacific, caused by the suppression of the ENSO signal (not shown).

5) Seasonal variations

The analysis of seasonal variations reveals that Fig. 1 merges precipitation changes of opposite sign that occur in different seasons. In fact, in Fig. 8, the mean precipitation for December through February (DJF) is compared with the precipitation for June through August (JJA). These overall patterns agree with the annual map (Fig. 1):

  • AL: Variations occur mostly on the tropical oceans, particularly over the Pacific; a strengthening of the Indian monsoon is evident.

  • AO: There is a general decrease in precipitation over continents and an increase over oceans.

  • A: There is a suggestion of nonlinearity in the responses, because the anomalies of AL and AO do not simply add together.

Of particular interest is the seasonal response relative to experiment AO. A dipolar pattern over tropical Africa in JJA suggests a southward shift of the ITCZ. An increase in DJF precipitation to the east of the southeastern United States suggests an increased baroclinic cyclogenetic activity in winter. Finally, an increase in JJA precipitation over most of the northern Atlantic suggests an overall strengthening of the northern Atlantic storm track in summer.

The analysis of the seasonal surface temperatures reveals that the AL run does not display any change, whereas the AO run shows a significant cooling of the Northern Hemisphere continents (not shown). In particular, there is a wintertime cooling over the eastern part of North America, consistent with increased surface temperature gradients toward the ocean, which in turn may lead to increased U.S. East Coast baroclinic cyclogenetic activity and precipitation (Fig. 8e). The general cooling over the Sahara, particularly in summer, is consistent with a weaker penetration of the ITCZ into the northern part of the African continent. The analysis of vorticity maps through different seasons is consistent with these statements. Particularly, a cyclonic vorticity anomaly is observed in run AO, in DJF, over the U.S. East Coast (not shown), consistently with the increased precipitation increases (Fig. 8e) and stronger surface temperature gradients. There is also a summer cyclonic vorticity anomaly over the northern Atlantic (not shown), suggesting a strengthening of the westerly flow and thus of the storm track over the eastern part of the northern Atlantic, in agreement with the precipitation anomaly (Fig. 8f).

Finally, the seasonal behavior of precipitation, surface temperatures, and vorticity for the A case confirm noticeable differences with the simple superpositions of the anomalies from the AL and AO cases (not shown), where the two sources of variability are separately disabled; for example, the two strong winter cyclonic anomalies in AL and AO over the eastern United States do not add in the A case, which, over that region, has a cyclonic anomaly that is weaker than the two other cases.

4. Changes in precipitation variability

a. Comparison with KA00

Figure 9 represents the variance of annual precipitation, area averaged over various regions of the world. Particularly, the histograms in Figs. 9a and 9d are averaged over all the land grid boxes between approximately 10°S and 10°N (tropical land) and all land gridboxes between approximately 30°N and 50°N (Northern Hemisphere midlatitude land), and can be compared with KA00 (their Fig. 2). In our study, however, we observe that important differences occur on the regional level, so other regions are also considered for this precipitation statistic.

For the strip of midlatitude land selected, there is a strong agreement between the control run produced by the COLA GCM coupled with SSiB (Fig. 9d) and the control run produced by KA00. Both cases display an interannual variance of approximately 1.9 × 104 mm2 yr−2. Agreement is also found analyzing the impact of suppressed SST variability (case AL): the variance is reduced by about 20% with respect to the control run in both cases. The impact of suppressed land variability (AO run) is more important than the impact of the ocean, in agreement with the KA00 study. However, our model is slightly more sensitive than the model used by KA00. In fact, the decrease in precipitation variance occurring in the AO run with respect to the control run is larger in our case than in the KA00 experiment. The smallest variability between the four experiments is obtained, as expected, by suppressing both the ocean and land sources of variability (experiment A). This finding is the same in our experiment as well as in KA00. However, the variance produced in our case A run is smaller than the one obtained in KA00's case A run. The variance ratios, which are the basis of the linear interpretation of results in KA00 study, are discussed in Part II of this article.

Our model and the model used by KA00 differ more substantially in the Tropics. The control runs still compare relatively well (both studies provide a variance in the range of 1.3–1.7 × 105 mm2 yr−2), but the proportion of land versus ocean variability is not in agreement. Our study suggests a larger reduction of precipitation variability due to suppression of land variability as compared to that induced by suppression of ocean variability (Fig. 9a). This result is opposite to that found by KA00 (their Fig. 2b). The reason is a greater sensitivity to continental evaporation in our model than in KA00's. In fact, the drop in precipitation variance induced by the mere suppression of interannually varying SSTs (run AL) is approximately the same in both experiments. Conversely, the suppression of land variability leads to a reduction in variance from 1.7 × 105 mm2 yr−2 (control) to 1.2 × 105 mm2 yr−2 (run AO) in the KA00 experiment, whereas in our case the variance changes from 1.3 × 105 to 0.3 × 105 mm2 yr−2.

In KS95, the suppression of land variability leads to a reduction in variance from 0.9 × 105 mm2 yr−2 (control) to 0.54 × 105 mm2 yr−2 (run AO). The difference between KS95 and KA00 was not caused by the choice of land surface scheme, which was the same in both studies, but was possibly due to stronger land–ocean teleconnections in the KA00 case.

b. Regional analysis

The response in precipitation variance that can be seen in Figs. 9a and 9d arises from an areal average that merges together very different local responses. The most striking example is given by the variances of mean annual precipitation over the Amazon (Fig. 9b) and Congo basins (Fig. 9c). First, the control variance of mean annual precipitation is almost twice as large for the Congo region (1.7 × 105 versus 0.9 × 105 mm2 yr−2). Second, the decrease in variance due to suppression of land variability is far more important for the Amazon than for the Congo region; this means that, in our model, the Amazon is more sensitive to land surface changes than the Congo region. This result is consistent with previous tropical deforestation studies (Polcher and Laval 1994; Sud et al. 1996).

Some discrepancies also occur between the continental United States and continental Eurasia. The latter region is chosen as a strip of land from eastern Europe to the Pacific, but having little extension in latitude, in order to cover most of the “transitional zone” between desert and taiga region. The variance analysis reveals that the variability of the control simulation in the U.S. case (Fig. 9e) is approximately 4 times larger than in the Eurasian case (Fig. 9f). Moreover, the impact of ocean over Eurasia is minimal: the region appears to be controlled mostly by land and by internal variability. In contrast, the ocean plays a relatively stronger role in controlling precipitation variance of the United States.

5. Conclusions

In this work we critically revisit the KS95 and KA00 experiments, reproducing the simulations with the COLA GCM, and find a general agreement in the precipitation variance diagnostics. However, the suppression of the interannual variations of land evaporability does not lead exclusively to changes in the variability of precipitation, but also causes important changes in the mean state of the atmosphere. The general circulation is affected, and so is the mean climate. Changes include a general decrease in the precipitation over land and an increase over ocean, a substantial cooling of the Northern Hemisphere land, a weakening of the Indian monsoon, a southward shift of the ITCZ, and a general change in the Northern Hemisphere midlatitude storm tracks. All these changes have been analyzed from a dynamical point of view, and they have been found to be consistent with the changes in the mean atmospheric circulation. Furthermore, we analyze the area-averaged variance of precipitation over the Tropics and the midlatitudes, comparing them with KA00 results. We find that our model is slightly more sensitive than the one adopted by KA00. This leads to a contrasting result for the variance of the mean annual precipitation over the Tropics: in our experiment the role of land is greater than the role of the ocean (KA00 found the opposite result). We also find important regional differences between the Congo and Amazon regions, the latter being much more sensitive to the land forcing than the former. A comparison of the precipitation response between Eurasia and northern America finds continental Eurasia generally less dominated than America by the ocean impact and more by internal atmospheric dynamics.

In Part II we address the hypothesis of linearity proposed by KA00 to explain the roles of sea and land in precipitation variability. In addition, we decompose the time structure of the land surface forcing by adding three more experiments, in which the interannual variability and diurnal cycle of evaporability are in turn enabled/disabled.

Acknowledgments

This work was supported by the omnibus grants from the National Science Foundation (ATM98-14295), the National Oceanic and Atmospheric Administration (NA96GP0056), and the National Aeronautics and Space Administration (NAG5-8202). We thank Dr. R. Koster for helpful comments and Dr. A. Schlosser for very valuable discussions.

REFERENCES

  • Bourke, W., McAvaney B. , Pury K. , and Thurling R. , 1977: Global modeling of atmospheric flow by spectral methods. Methods in Computational Physics, Vol. 17, J. Chang, Ed., Academic Press, 267–324.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., 1991: Mid-latitude Weather Systems. Harper Collins, 507 pp.

  • Davies, R., 1982: Documentation of the solar radiation parameterization in the GLAS climate model. NASA Tech. Memo. 83961, 57 pp.

  • DeWitt, D. G., 1996: The effect of the cumulus convection scheme on the climate of the COLA general circulation model. COLA Tech. Rep. 27, 58 pp. [Available from COLA, 4041 Powder Mill Rd., Suite 302, Calverton, MD 20705.].

    • Search Google Scholar
    • Export Citation
  • Dirmeyer, P. A., and Zeng F. J. , 1999: An update to the distribution and treatment of vegetation and soil properties in SSiB. COLA Tech. Rep. 78, 25 pp. [Available from COLA, 4041 Powder Mill Rd., Suite 302, Calverton, MD 20705.].

    • Search Google Scholar
    • Export Citation
  • Fennessy, M. J., and Coauthors. 1994: The Simulated Indian Monsoon: A GCM sensitivity study. J. Climate, 7 , 3343.

  • Harshvardhan, R., Davies R. , Randall D. A. , and Corsetti T. G. , 1987: A fast radiation parametrization for atmospheric circulation models. J. Geophys. Res., 92 ((D1),) 10091016.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hastenrath, S., 1985: Climate and Circulation of the Tropics. Kluwer Academic, 455 pp.

  • Kasahara, A., 1974: Various vertical coordinate systems used for numerical weather prediction. Mon. Wea. Rev., 102 , 509522.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kiehl, J. T., Hack J. J. , Bonan G. , Boville B. A. , Williamson D. L. , and Rasch P. J. , 1998: The National Center for Atmospheric Research Community Climate Model: CCM3. J. Climate, 11 , 11311149.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kinter, J. L., Shukla J. , Marx L. , and Schneider E. K. , 1988: A simulation of the winter and summer circulations with the NMC global circulation model. J. Atmos. Sci., 45 , 24862522.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kinter, J. L., and Coauthors. 1997: The COLA atmosphere–biosphere general circulation model. Volume 1: Formulation. COLA Tech. Rep. 51, 1–44. [Available from COLA, 4041 Powder Mill Rd., Suite 302, Calverton, MD 20705.].

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Suarez M. J. , 1995: Relative contributions of land and ocean processes to precipitation variability. J. Geophys. Res., 100 , 1377513790.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Heiser M. , 2000: Variance and predictability of precipitation at seasonal-to-interannual timescales. J. Hydrometeor., 1 , 2646.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lacis, A. A., and Hansen J. E. , 1974: A parameterization for the absorption of solar radiation in the earth's atmosphere. J. Atmos. Sci., 31 , 118133.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mellor, G. L., and Yamada T. , 1982: Development of a turbulence closure model for geophysical fluid processes. Rev. Geophys. Space Phys., 20 , 851875.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miyakoda, K., and Sirutis J. , 1977: Comparative integrations of global spectral models with various parametrized processes of subgrid scale vertical transports. Beitr. Phys. Atmos., 50 , 445447.

    • Search Google Scholar
    • Export Citation
  • Moorthi, S., and Suarez M. J. , 1992: Relaxed Arakawa–Schubert: A parameterization of moist convection for general circulation models. Mon. Wea. Rev., 120 , 9781002.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Palmer, T. N., Shutts G. J. , and Swinbank R. , 1986: Alleviation of a systematic westerly bias in general circulation and numerical weather prediction models through an orographic gravity wave parametrization. Quart. J. Roy. Meteor. Soc., 112 , 10011039.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Polcher, J., and Laval K. , 1994: A statistical study of the regional impact of deforestation on climate in the LMD GCM. Climate Dyn., 10 , 205219.

  • Reale, O., Dirmeyer P. , and Schlosser A. , 2002: Modeling the effect of land surface evaporation variability on precipitation variability. Part II: Time- and space-scale structure. J. Hydrometeor.,3, 451–466.

    • Search Google Scholar
    • Export Citation
  • Reynolds, R. W., and Smith T. M. , 1994: Improved global sea surface temperature analyses using optimal interpolation. J. Climate, 7 , 929948.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sellers, P. J., Mintz Y. , Sud Y. C. , and Dalcher A. , 1986: A Simple Biosphere Model (SiB) for use within general circulation models. J. Atmos. Sci., 43 , 505531.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Simmons, A. J., and Struefing R. , 1981: An energy and angular-momentum conserving finite-difference scheme, hybrid coordinates and medium-range weather prediction. ECMWF Tech. Rep. 28, 68 pp. [Available from ECMWF, Shinfield Park, Reading RG2 9AX, United Kingdom.].

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., Walker G. K. , Kim J-H. , Liston G. E. , Sellers P. J. , and Lau W. K-M. , 1996: Biogeophysical consequences of a tropical deforestation scenario: A GCM simulation study. J. Climate, 9 , 32253247.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tiedtke, M., 1984: The effect of penetrative cumulus convection on the large-scale flow in a general circulation model. Beitr. Phys. Atmos., 57 , 216239.

    • Search Google Scholar
    • Export Citation
  • Xue, Y-K., Sellers P. J. , Kinter J. L. , and Shukla J. , 1991: A simplified biosphere model for global climate studies. J. Climate, 4 , 345364.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xue, Y-K., Zeng F. J. , and Schlosser C. A. , 1996: SSiB and its sensitivity to soil properties. A case study using HAPEX–Mobilhy data. Global Planet. Change, 13 , 183194.

    • Crossref
    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

(a) Mean precipitation (mm day−1) for the control run. The mean is calculated from 48 yr of integration (1951–98). All the subsequent means are computed with the same time frame. (b) AL mean precipitation departure from the control. (c) AO mean precipitation departure from the control run. (d) Simulation A mean precipitation departure from the control run. The solid (dotted) line represents areas where the increase (decrease) in precipitation is statistically significant.

Citation: Journal of Hydrometeorology 3, 4; 10.1175/1525-7541(2002)003<0433:MTEOLS>2.0.CO;2

Fig. 1.
Fig. 2.
Fig. 2.

(a) Mean evaporation (mm day−1) for the control run. (b) AL mean evaporation departure from the control run. (c) AO mean evaporation departure from the control run. (d) Simulation A mean evaporation departure from the control run. The solid (dotted) line represents areas where the increase (decrease) in evaporation is statistically significant

Citation: Journal of Hydrometeorology 3, 4; 10.1175/1525-7541(2002)003<0433:MTEOLS>2.0.CO;2

Fig. 2.
Fig. 3.
Fig. 3.

(a) Mean surface temperature (°C) for the control run. (b) AO mean surface temperature departure from the control run

Citation: Journal of Hydrometeorology 3, 4; 10.1175/1525-7541(2002)003<0433:MTEOLS>2.0.CO;2

Fig. 4.
Fig. 4.

(a) Mean divergence at σ = 0.9205 (s−1) for the control run. (b) AL mean divergence departure from the control run. (c) AO mean divergence departure from the control run. (d) Simulation A mean divergence departure from the control run

Citation: Journal of Hydrometeorology 3, 4; 10.1175/1525-7541(2002)003<0433:MTEOLS>2.0.CO;2

Fig. 4.
Fig. 5.
Fig. 5.

(a) Mean divergence at σ = 0.275 (s−1) for the control run. (b) AL mean divergence departure from the control run. (c) AO mean divergence departure from the control run. (d) Simulation A mean divergence departure from the control run

Citation: Journal of Hydrometeorology 3, 4; 10.1175/1525-7541(2002)003<0433:MTEOLS>2.0.CO;2

Fig. 5.
Fig. 6.
Fig. 6.

(a) Mean relative vorticity at σ = 0.5 (s−1) for the control run. (b) AO mean vorticity departure from the control run. Cyclonic vorticity is light (dark) in the Northern (Southern) Hemisphere

Citation: Journal of Hydrometeorology 3, 4; 10.1175/1525-7541(2002)003<0433:MTEOLS>2.0.CO;2

Fig. 7.
Fig. 7.

(top) Mean vertical velocity (10−3 hPa s−1) between 22.5°W and 15°E: (left) control run, (middle) AO, and (right) AO departure from the control run. Contour intervals are every 0.01 × 10−3 hPa s−1 (control and AO) and every 0.0025 × 10−3 hPa s−1 (AO minus control), with the zero contour omitted. (bottom) Meridional wind (m s−1): (left) control run, (middle) AO, and (right) AO departure from the control run. Contour intervals are every m s−1 (control and AO) and every 0.25 s−1 (AO minus control), with the zero contour omitted

Citation: Journal of Hydrometeorology 3, 4; 10.1175/1525-7541(2002)003<0433:MTEOLS>2.0.CO;2

Fig. 8.
Fig. 8.

Mean precipitation (mm day−1) for (left) DJF and (right) JJA: (a), (b) control run; (c), (d) AL mean precipitation departure from the control run; (e), (f) AO mean precipitation departure from the control run; (g), (h) simulation A mean precipitation departure from the control run

Citation: Journal of Hydrometeorology 3, 4; 10.1175/1525-7541(2002)003<0433:MTEOLS>2.0.CO;2

Fig. 9.
Fig. 9.

Histogram with the variance of annual precipitation for each experiment: (a) tropical land (11.35°S–11.35°N); (b) Amazon basin (11.35°S–0°, 74.625°–50.625°W); (c) Congo basin (7.425°S–3.715°N, 13.15°–31.875°E); (d) Northern Hemisphere midlatitude land (29.685°–51.955°N); (e) continental United States to the east of the Rockies (29.685°–51.955°N, 110.625°–88.125°W); and (f) continental Eurasia (44.535°–63.085°N, 13.125°–136.875°E)

Citation: Journal of Hydrometeorology 3, 4; 10.1175/1525-7541(2002)003<0433:MTEOLS>2.0.CO;2

Save